Hello everybody.
I am in search of advice from moderators, as someone who makes third-party audio edits which I know are on a no-mans-land on this site, understandably.

None of my uploads have survived, and after the first few fuckups I began uploading them with the proper encoding and the characteristics of the original video.

So my question is, is there a checklist or something to keep in mind to -correctly- upload this whitout getting deleted?
I know for a fact in an old post from Mairo that "...if it's third party edit, it's closer to being deleted than approved [...] That's why sound edits shouldn't be accepted to begin with because it's third party edit that's identical to already exsisting version, but there has been given some leeway to sound edits."

What would be the requirements for me to make the most correct use possible of this "Possible leeway"?

I really don't want to depend on "x" to share my stuff X.X

Edits need to be the same quality as or better than the original. So if an edit is lower resolution, visibly more compressed, etc when compared to the original, it will be deleted. For the best chances, you should take our version of the original and put your audio on top of that. If you're doing something like downloading the original off of twitter, that will almost always be worse than what we have, which will result in the edit being deleted for being lower quality than the original. Since you aren't (or at least shouldn't be) touching the visuals, you should just be able to lay the audio track right on top of the original video we have.

For an actual post example, you uploaded post #4092004 which is an edit of post #3941285. If you look to the left where the file information is, there's a stark difference in filesize. The original is 53MB, meanwhile your edit is only 5MB. That large filesize on the original isn't just there for fun, that edit significantly reduced the quality. Exactly what and/or where isn't going to be clear, but something definitely happened. That makes the edit worse than the original, thus not allowed.

Are you familiar with ffmpeg to know if this code from ffmpeg meets the "don't touch the video" criteria?

#ffmpeg -i ORIGINALDOWNLOAD.mp4 -i MYAUDIOTRACK.wav -c copy OUTPUTVIDEO.mp4

Also, I would assume with this info that adding any text to the video like "original by: x audio by: x" or a waterwark would be a no go, yes?

reveriexs said:
Are you familiar with ffmpeg to know if this code from ffmpeg meets the "don't touch the video" criteria?

#ffmpeg -i ORIGINALDOWNLOAD.mp4 -i MYAUDIOTRACK.wav -c copy OUTPUTVIDEO.mp4

In theory yes, that should not touch the video. If you're doing this and seeing an issue, the problem is likely coming from when you're converting that MP4 to a WEBM, the quality drop is happening there.

reveriexs said:
Also, I would assume with this info that adding any text to the video like "original by: x audio by: x" or a waterwark would be a no go, yes?

I'm unsure of exact specifics as I am not a Janitor myself, but I do believe a watermark should be fine. The problem comes from reencoding the entire video to add that watermark, which usually leads to a drop in quality.

reveriexs said:
Are you familiar with ffmpeg to know if this code from ffmpeg meets the "don't touch the video" criteria?

#ffmpeg -i ORIGINALDOWNLOAD.mp4 -i MYAUDIOTRACK.wav -c copy OUTPUTVIDEO.mp4

Also, I would assume with this info that adding any text to the video like "original by: x audio by: x" or a waterwark would be a no go, yes?

Except MP4 formats are not supported on e621 and you shouldn't really have WAV audio track in there. Also if the MP4 does have audio track, even emtpy one, it might not map all the inputs to correct outputs and those might need to be specified. Also if the audio or video is already in correct codec, you can specify to only copy those and then transcode the other tracks.
Adding watermarks and transcoding are fine (and you have to encode if you are adding visuals like watermarks or making the video longer or altering speed, etc. etc.) as long as the quality is "transparent", as in the quality is similar to source material.

mairo said:
Except MP4 formats are not supported on e621 and you shouldn't really have WAV audio track in there. Also if the MP4 does have audio track, even emtpy one, it might not map all the inputs to correct outputs and those might need to be specified. Also if the audio or video is already in correct codec, you can specify to only copy those and then transcode the other tracks.
Adding watermarks and transcoding are fine (and you have to encode if you are adding visuals like watermarks or making the video longer or altering speed, etc. etc.) as long as the quality is "transparent", as in the quality is similar to source material.

I'll definitively will have to study and understand this before trying again. Video enconding and editing is new to me.

Which is the codec used on e621? Or just the container is specified?
But yes, the videos downloaded from the site are on webm container and for what i found the "appropiate" audio format then would be .ogg?
So could I use the same code above but replacing mp4 with webm and wav with ogg?